zurichopenrepositoryand year: 2016 · firmicutes increased when compared to the non-enriched mung...
Post on 26-Aug-2020
3 Views
Preview:
TRANSCRIPT
Zurich Open Repository andArchiveUniversity of ZurichMain LibraryStrickhofstrasse 39CH-8057 Zurichwww.zora.uzh.ch
Year: 2016
Effects of different media on the enrichment of low numbers of Shigatoxin-producing Escherichia coli in mung bean sprouts and on the
development of the sprout microbiome
Margot, Heike ; Tasara, Taurai ; Zwietering, M H ; Joosten, Han ; Stephan, Roger
Abstract: Sprouted seeds have been implicated in a number of serious outbreaks caused by Salmonellaand Shiga toxin-producing Escherichia coli. Sprouts pose a very complex challenge to bacterial pathogenenrichment and detection since they naturally contain high levels of background microflora includingmembers of the Enterobacteriaceae. As such, the currently used method cannot ensure reliable detectionof STEC in sprouts. In this study, we compared different media for the enrichment of Enterobacteriaceaein their ability to promote the growth of stressed STEC at 37°C and 42°C. Mung bean sprouts werespiked with low levels of STEC and their growth was recorded over time. In addition, the microbiomeof mung bean sprouts was analysed before and after enrichment. Our results indicate that the growth ofdry-stressed STEC is comparable in all of the tested enrichment media except for mTSB+Novobiocin andnot influenced by the incubation temperature. Low levels of STEC spiked into the sprouts resuspendedin media only grew to levels of around 4logcfu/ml during enrichment, which could reduce the probabilityof detection. Proteobacteria was the dominant phylum detected within the microbiome of non-enrichedmung bean sprouts. During enrichment in EE-broth, Proteobacteria remained the most abundant phy-lum. In contrast, during enrichment in BPW the relative abundance of Proteobacteria decreased whereasFirmicutes increased when compared to the non-enriched mung bean sprout microbiome. The micro-biome composition was not significantly influenced by the incubation temperature during enrichment inboth BPW and EE-broth. This is the first study to examine the microbiome on sprouted mung beanseeds during BPW and EE enrichment and relates the bacterial community composition changes to theenrichment of pathogens.
DOI: https://doi.org/10.1016/j.ijfoodmicro.2016.05.005
Posted at the Zurich Open Repository and Archive, University of ZurichZORA URL: https://doi.org/10.5167/uzh-130007Journal ArticleAccepted Version
The following work is licensed under a Creative Commons: Attribution-NonCommercial-NoDerivatives4.0 International (CC BY-NC-ND 4.0) License.
Originally published at:
Margot, Heike; Tasara, Taurai; Zwietering, M H; Joosten, Han; Stephan, Roger (2016). Effects of differentmedia on the enrichment of low numbers of Shiga toxin-producing Escherichia coli in mung bean sproutsand on the development of the sprout microbiome. International Journal of Food Microbiology, 232:26-34.DOI: https://doi.org/10.1016/j.ijfoodmicro.2016.05.005
2
1
1
2
Evaluation of different buffered peptone water (BPW) based enrichment broths 3
for detection of Gram-negative foodborne pathogens from various food matrices 4
5
6
H. Margot1, M.H. Zwietering
2, H. Joosten
3, Emer O`Mahony
4, R. Stephan
1 7
8
1Institute for Food Safety and Hygiene, Vetsuisse Faculty University of Zurich, CH-8057 9
Zurich, Switzerland 10
2Laboratory of Food Microbiology, Wageningen University, 6700 AA Wageningen, the 11
Netherlands 12
3European Chair in Food Safety Microbiology, Wageningen University, 6700 AA 13
Wageningen, the Netherlands 14
4Quality and Safety Department, Nestlé Research Center, Vers-chez-les-Blanc, CH-1000 15
Lausanne, Switzerland 16
17
18
19
20
21
22
23
* Corresponding author: 24
R. Stephan, Institute for Food Safety and Hygiene, Vetsuisse Faculty University of Zurich, 25
Winterthurerstr. 272, CH-8057 Zurich, Switzerland, Phone 0041-44-6358651, fax 0041-44-26
6358908, e-mail stephanr@fsafety.uzh.ch 27
2
Abstract (max 400 words) 28
This study evaluated the effects of changing the composition of the pre-enrichment medium 29
buffered peptone water (BPW) on the growth of stressed and unstressed Gram-negative 30
foodborne pathogens in a one-broth enrichment strategy. BPW supplemented with an 31
available iron source and sodium pyruvate, along with low levels of 8- hydroxyquinoline and 32
sodium deoxycholate (BPW-S) improved the recovery of desiccated Cronobacter spp. from 33
powdered infant formula. Growth of Salmonella and STEC was comparable in all BPW 34
variants tested for different food matrices. In products with high levels of Gram-negative 35
background flora (e.g. sprouts), the target organisms could not be reliably detected by PCR in 36
any of the BPW variants tested unless the initial level exceeded 103 cfu/10g of sprouts. 37
Based on these results we suggest BPW-S for a one-broth enrichment strategy of stressed 38
Gram-negative foodborne pathogens from dry products. However, a one-broth enrichment 39
strategy based on BPW variants tested in this evaluation is not recommended for produce 40
with a high level of Gram-negative background flora due to very high detection limits. 41
42
43
44
45
46
47
Keywords: pre-enrichment, one-broth strategy, supplements, resuscitation, Salmonella, 48
STEC, Cronobacter, Enterobacteriaceae 49
3
1. Introduction 50
Key to the reliability of all methods for the detection of Gram-negative foodborne pathogens 51
(e.g Salmonella, Cronobacter and Shigatoxin-producing E. coli) is the capability to recover 52
low numbers of stressed cells from any kind of food matrix to a detectable level. These 53
methods usually consist of three consecutive steps. The pre-enrichment step, that is common 54
to most current detection methodologies (cultural, molecular and immunological), aims at the 55
resuscitation/ recovery of cells in a non-selective medium such as buffered peptone water 56
(BPW). The second step, a selective enrichment, should promote growth of the recovered 57
target bacteria. The last step, the actual detection of the pathogen can be accomplished with 58
either cultural or molecular methods. 59
Gram-negative foodborne pathogens are usually present in relatively low numbers, sometimes 60
accompanied by a high number of closely related competitor organisms (Baylis et al., 2000). 61
In addition, cells may be sub lethally injured by food processing or intrinsic factors of the 62
food matrix (Edel and Kampelmacher, 1973). The purpose of the pre-enrichment step is to 63
allow stressed target microorganisms to resuscitate and grow in either a non-selective or 64
moderately selective environment. At this stage, the target organisms can be overgrown by 65
the background flora in the food product due to absent selectivity of the medium, which can 66
lead to false negative results. 67
The selective enrichment should promote growth of the recovered target bacteria. The 68
necessity of applying a selective enrichment will mostly depend on the characteristics of the 69
food product, particularly on its microbial flora. In products with a high number of competing 70
organisms, in which the target bacteria are likely to be stressed, both a non-selective and 71
selective enrichment step will be necessary. For products with low levels of background 72
flora, a one-broth enrichment strategy in an unselective enrichment broth can be sufficient. 73
Since the pre-enrichment and selective enrichment step are the main time limiting factors in 74
4
regards to turn around time for rapid methods (Baylis et al., 2000), a moderately selective 75
one-broth enrichment strategy can decrease the total time until results are available. 76
The success of the pre- enrichment step is however not only depending on the food product 77
but also on the quality of the medium. Preliminary work has shown that there are some 78
inconsistencies in the performance of BPW and that the ISO standard formulation results in 79
performance variations not just from manufacturer to manufacturer but also from batch to 80
batch. A comparison of 18 different commercially brands of buffered peptone water showed 81
differences in performance regarding buffering capacity, growth of unstressed bacteria and 82
recovery of dry-stressed cells (unpublished data). The performance of the BPW brands 83
depended heavily on the matrix and the organism tested. Baylis et al. (2000) compared two 84
commercially available preparations of buffered peptone water with regards to their ability to 85
promote recovery and growth of sub-lethally injured S. enterica. There was a significantly 86
higher recovery rate stressed of cells with one brand of buffered peptone water however, the 87
performance of BPW in food was mostly dependent on the food type with the type and level 88
of background flora present being the most important influential factors. Strain variability 89
also exerted an influence on the recovery of stressed cells. Different studies have investigated 90
an improvement of enrichment media by supplementation and by modification of the 91
conditions during enrichment such as temperature, pH, time etc. (Andrews et al., 1986). Gray 92
et al. (2006) investigated the effect of variation between classes of casein, gelatine and 93
peptones from different sources. They concluded that the variation in peptone has a 94
substantial influence when looking at growth and enumeration of bacteria. For STEC, pre-95
enrichment methods were developed based on meat industry requirements that are, however, 96
inadequate for the recovery of STEC from vegetables, flour and other low aw matrices (Sata et 97
al., 2003). Weber et al. (2009) showed that the supplementation of growth factors such as 98
additional iron and sodium pyruvate, along with low levels of inhibiting agents primarily 99
5
against Gram-positive background flora (8- hydroxyquinoline and sodium deoxycholate), 100
enhances the recovery rate of stressed cells. 101
Products containing starter cultures and probiotic cultures present a special challenge to 102
pathogen detection. Strong acidification of the enrichment caused by lactic acid producing 103
bacteria can lead to inactivation of target bacteria and subsequently false-negative results. 104
Previous experimental work has shown that increased concentrations of buffer phosphates in 105
BPW improved detection of pathogens in powders containing probiotics (unpublished data). 106
For this reason, BPW with 3x and 6x increase in buffer phosphate concentration were 107
included in the study. So far, there are only few studies evaluating the use of BPW with 108
supplements in a one-broth enrichment. In particular, data obtained from experiments in the 109
food matrix are scarce. 110
The aim of the current study was to compare different modifications of BPW with regard to 111
their ability to promote growth of unstressed and stressed Gram-negative foodborne 112
pathogens in different food matrices. 113
114
2. Materials and methods 115
2.1. Bacterial strains 116
Different Salmonella enterica, Cronobacter spp. and Shigatoxin-producing Escherichia coli 117
(STEC) strains were chosen as representatives for Gram-negative food borne pathogens 118
(Table 1). Working cultures were made from frozen (-80°C) BHI (Oxoid CM1135, 119
Basingstoke, United Kingdom) with 20% glycerol stocks and maintained on blood agar plates 120
(Difco Columbia blood agar base, 5% sheep blood, CM0031 Oxoid) at 4± 1°C. All strains 121
were natural isolates, obtained from our in-house collection. 122
123
2.2 Supplementation of BPW 124
6
Buffered peptone water (BPW, Oxoid CM1049) was prepared according to the 125
manufacturer’s instructions. BPW-S was prepared by using BPW with the following 126
supplements added before autoclaving: 40 µM 8- hydroxyquinoline, 0.5 g/l ammonium-127
iron(III) citrate, 0.1 g/l sodium deoxycholate, 0.1g /l sodium pyruvate (all from Sigma-128
Aldrich, Buchs, Switzerland). 129
6xBPW (quantities for 3xBPW in brackets) was prepared by addition of 7.5 g/l (3 g/l) 130
KH2PO4 and 17.5 g/l (7 g/l) Na2HPO4 (anhydrous) (Sigma-Aldrich) to BPW. 6xBPW-S 131
contains both additional buffer salts and supplements of BPW-S. Throughout this work, 132
buffered peptone water from a single batch was used to avoid lot-to-lot variations. 133
134
2.3 Recovery of dry stressed cells 135
Each ten strains of Cronobacter spp., STEC and S. enterica were grown overnight in 9 ml 136
BHI (Oxoid) at 37°C. Grown cultures were centrifuged at 13000 g for 10 min. The 137
supernatant was discarded and the pellet was re-suspended in 1 ml BPW. The cell suspension 138
with cell counts of approximately 109-10
10 cfu/ml was serially tenfold diluted in BPW. 10 µl 139
of each dilution were pipetted in eight wells of a 96- well microtiter plate using one plate per 140
strain resulting in 107- 10
8 cfu/well in the first row. Cell counts of the culture were determined 141
with plate counts on tryptic soy agar (BD Diagnostic Systems, Heidelberg, Germany). After 142
two days of storage in a desiccator containing silica gel at room temperature, one complete 143
plate was rehydrated with 200 µl BPW, BPW-S or 3xBPW per well and incubated 16± 2 h at 144
37°C. The most probable number (MPN) of microorganisms that had survived the drying 145
process and were able to grow in the enrichment broth was determined based on the number 146
of wells in which growth was observed using the Bacteriological Analytical Manual Online 147
MPN table. 148
(http://www.fda.gov/Food/FoodScienceResearch/LaboratoryMethods/ucm109656.htm) 149
150
7
2.4 Application of desiccation stress 151
In food matrices with a low aw, pathogens are expected to have encountered desiccation 152
stress. To achieve a more realistic scenario, cells used for spiking were desiccation stressed 153
before inoculation. 154
The STEC K124 and 33, Cronobacter spp. E615 and E776, S. enterica N10 1905 and N472 155
962 strains were grown separately overnight in 30 ml BHI (Oxoid) at 37°C. After 156
centrifugation for 12 min at 4000 rpm and the supernatant was discarded. The inside of the 157
tube was dried with a sterile cotton swab. Small amounts of sterile CaCO3 (Sigma-Aldrich) 158
were added to the pellet and mixed vigorously with a spatula until homogenously distributed. 159
More CaCO3 was added to a final amount of approximately 15 g of powder. The powder was 160
stored in a desiccator with silica gel for two days. The homogeneity of the powder was 161
verified one day before the actual experiment by weighing five times 0.1 g of powder, 162
diluting in 0.9% saline and plating on selective agar (Rapid` E. coli 2 agar for E. coli (Biorad, 163
Marnes la Coquette, France), X.L.D agar for S. enterica (Oxoid), Brilliance Enterobacter 164
sakazakii agar for Cronobacter spp. (Oxoid)). However, plating on selective agar might have 165
resulted in lower cell counts than those actually found in the powder. 166
To obtain the desired concentration, 0.1 g of the powder with bacteria was suspended in 0.9% 167
saline, diluted if necessary and added to the enrichment. The spiking levels were checked by 168
enumeration on selective agar. 169
170
2.5 Growth of pathogens inoculated into food enrichments 171
10 g samples of skim milk powder, oat flakes, minced beef, soy sprouts and powdered infant 172
formula (PIF) were diluted 1/10 with the different enrichment media and mixed in a 173
stomacher for 30 s at medium intensity. Food products for one experiment were obtained 174
from the same supplier, if possible originating from the same batch. The samples were 175
inoculated with low levels of the target strain (aiming at 101 cfu in 100 ml) and incubated for 176
8
18±2 h at 37°C. For spiking with cells without application of stress (for minced beef and 177
sprouts), an overnight culture grown in BHI was diluted to the desired concentration and 178
added to the diluted matrix. In the case of milk powder, oat flakes and PIF, the target strains 179
were desiccation stressed before inoculation, using the protocol described in 2.4. Skimmed 180
milk powder was inoculated with Salmonella, oat flakes were inoculated with STEC and 181
Salmonella, minced beef was inoculated with STEC, soy sprouts were inoculated with STEC 182
and S. enterica and powdered infant formula was inoculated with Cronobacter spp. Inoculum 183
levels were confirmed by plate counts on tryptic soy agar. After 4, 8, 15, 20 and 24 h of 184
incubation, samples of each enrichment broth were diluted in 0.9 % saline and plated on the 185
appropriate selective agar. To verify that the used food products were not naturally 186
contaminated with S. enterica, Cronobacter spp. or STEC, negative controls were incubated 187
and checked for the presence of pathogens by plating on selective agar or by Assurance GDS 188
real-time PCR (Biocontrol, Bellevue, USA). Total aerobic counts and E. coli/ coliform counts 189
were determined in the un-inoculated samples by plating them on tryptic soy agar at regular 190
intervals. Minced meat and sprout samples were additionally plated on Rapid` E. coli 2 agar 191
(Biorad). All agar plates were aerobically incubated at 37°C for 24±2 h. The pH of the 192
negative control was measured at regular intervals in 6xBPW and BPW-S using a pH meter 193
(Orion Star, Thermofisher Scientific, Reinach, Switzerland). To obtain data after more then 194
12 h incubation an identical second pre-enrichment broth was prepared in the late afternoon, 195
also incubated at 37°C, and sampled the next day. This approach may result in an increase of 196
the variability. 197
198
2.6 Recovery of STEC and Salmonella from sprouts 199
Fresh soy sprouts were obtained from a local Asian supermarket. 10 g samples were diluted 200
1/10 with BPW, BPW-S and 6xBPW and mixed in a Stomacher for 30 s at medium intensity. 201
Sprouts were spiked with 10, 102, 10
3 cfu / sample with E. coli (strain K124) and a naturally 202
9
nalidixic acid resistant Salmonella Kentucky (strain N08 2487) and incubated for 18±2 h at 203
37°C. Plate counts were performed to confirm the inoculation level. After incubation, the 204
samples were diluted and plated on selective media. Samples inoculated with STEC were 205
plated on Rapid` E. coli 2 agar (Biorad) and samples inoculated with S. enterica were plated 206
on Luria Bertani Agar + 256 µg/ml nalidixic acid sodium salt (Sigma-Aldrich) Plates were 207
incubated for 24 h at 37°C. In addition, presence of the pathogens was tested with the 208
Assurance GDS real-time PCR. An un-inoculated sample was used as a negative control. 209
210
2.7 Statistical analysis 211
Statistical analysis was performed using IBM SPSS. The effect of the enrichment media on 212
the resuscitation of desiccation stressed cells was compared applying one-way ANOVA. 213
A three-way ANOVA was used to compare the cell counts from the enriched cultures for the 214
four enrichment media, five food matrices and the four organisms. 215
216
3. Results and Discussion 217
3.1. Recovery of desiccated cells 218
The MPN concentrations (log cfu/ml) after rehydration and incubation for the tested strains 219
using the three enrichment media are shown in Figure 1. Salmonella strains were recovered to 220
values reaching from 0.5 log MPN to 4.6 log MPN. The mean MPN counts were 3.6± 1.2, 221
3.7± 1.3 and 3.4± 1.3 in BPW, BPW-S and 6xBPW, respectively. Desiccated Salmonella 222
strains did not show great variation in regards to the different media. Two strains remained at 223
low cfu levels after rehydration and incubation, which indicates that growth after desiccation 224
was more strain than media dependent. Desiccated STEC strains were recovered to highest 225
concentrations in BPW (mean MPN count 4.1± 0.2), showing little strain variation. The mean 226
concentrations in BPW-S and 3xBPW were up to 2.0 log MPN and 0.6 log MPN lower, 227
respectively. There was no statistical significant difference between the media (P= 0.87). 228
10
Highest concentrations for Cronobacter strains were recorded in BPW-S (mean MPN count 229
2.5± 0.7) and 6xBPW (mean MPN count 3.3± 0.9) (P= 0.06). However, strain variability 230
resulted in differences in counts of up to 2.5 log MPN. The strains showed least resuscitation 231
in BPW (mean MPN count 3.2± 0.5). 232
The positive effect of the combined supplementation of BPW with an iron source, sodium 233
pyruvate and sodium deoxycholate on stressed Enterobacteriaceae was already described by 234
Weber et al. (2009). Nevertheless, recovery of desiccated STEC was significantly superior in 235
BPW without supplements (P< 0.01). One of the supplements, sodium deoxycholate, is the 236
selective component in bile salts and its use in culture media for the inhibition of Gram-237
positive organisms and selection of Gram-negatives is well recognized (MacConkey, 1905). 238
However, the negative effect of bile on the detection of stressed pathogens has also been 239
described (D`Mello et al., (1987); Hara-Kudo et al. (2000). A concentration of 1.0 µg/ml 240
sodiumdeoxycholate in an agar medium resulted in increased lag time and growth rate of E. 241
coli and the interference with the flagellation was named as a possible reason by D`Mello et 242
al. (1987). Stephens and Joynson (1998) reported decreased recovery of acid/salt stressed E. 243
coli O157:H7 in tryptone soya broth when bile salts were added. These findings indicate that 244
bile salts with sodium deoxycholate being the main ingredient can deteriorate the recovery of 245
stressed E. coli cells (Stephens and Joynson, 1998). 246
Sodium pyruvate protects stressed bacteria from reactive oxygen species (MacDonald et al., 247
1983). These can develop in the enrichment broth through auto-oxidation of reducing sugars 248
and cause decreased recovery of stressed cells (Stephens et al., 2000). Sodium pyruvate is 249
able to facilitate the repair process in injured bacteria and can replace the function of catalase 250
in these organisms. No negative effects on any of the organisms tested in our study are 251
described in literature. Therefore the observed inhibitory effect in STEC is most probably not 252
caused by this substance. 8-hydroxyquinoline (8-HQ) was added to the enrichment medium to 253
11
replace vancomycin as an inhibitor for Gram-positive competitors as this compound is heat 254
stable and alternatives to antibiotics are desirable. (Weber et al., 2009). 255
256
257
3.2. Maximum population density in the matrix 258
The maximum population density for each matrix with standard deviation is shown in Figure 259
2. All food matrices were inoculated with two strains of the same species and the experiments 260
were performed in triplicate. In one experiment, one food product was spiked with one strain 261
and diluted with all tested media. The spiking inoculum was identical for one experiment. 262
However, spiking levels between different replicates can differ which is the main reason for 263
the high standard deviation in some cases. When PIF was spiked with low levels of 264
desiccation stressed Cronobacter, highest counts were seen in BPW-S in all experiments. 265
Levels in the other enrichment media were on average 1 log cfu/ml lower than in BPW-S. 266
However, in two of the experiments, counts in 6xBPW and 6xBPW-S were as high as in 267
BPW-S. Overall, the cell counts of Cronobacter were not significantly different between the 268
media (P= 0.06). Weber et al. (2009) demonstrated the same improved recovery with ten 269
Cronobacter spp. strains. They also showed that isolation from other naturally contaminated 270
samples of different origin was enhanced 271
For skimmed milk powder, Salmonella strain N472 962 grew to 9 log cfu/ml in all 272
enrichment media. Variation was small in this case (mean standard deviation= 0.5). Strain 273
N10 905 grew to lower counts in all media and experiments were yielding higher standard 274
deviations (mean standard deviation= 1.3). 275
STEC strains 33 and K124 spiked into cereals grew to 9 log cfu/ml, independent of the 276
enrichment medium. In contrast to the desiccation experiments (see 3.1.), both STEC strains 277
grew to same concentration in BPW and BPW-S. When S. enterica was spiked into cereals, 278
12
growth of the target bacteria again was lower in 6xBPW and 6xBPW-S than in BPW and 279
BPW-S with high standard deviation. 280
Minced meat was spiked with low levels of unstressed bacteria. STEC strain 33 grew to levels 281
of 8 log cfu/ml to 10 log cfu/ml in BPW-S, 6xBPW and 6xBPW-S. Values in BPW were up 282
to 2 log cfu/ml lower. Minced meat was also spiked with S. enterica. For both strains, all four 283
media showed the same maximum population density. 284
The general outcome of the matrix experiments is that both Salmonella and STEC strains 285
(with the exemption of K124) did grow to comparable concentrations in all four enrichment 286
media independent of the matrix. This indicates that the food matrix balances the effects of 287
the media supplements, both positive and negative. The results also show that in the matrices 288
tested, Salmonella, STEC and Cronobacter spp. can be isolated with a one-broth strategy with 289
BPW-S. In food products in which one or more of these three targets have to be tested, one 290
enrichment could be sufficient and omission of a selective enrichment allows time saving. 291
6xBPW was included in the study because previous experimental work showed that increased 292
buffering capacity enhanced recovery of desiccated cells from products with a high number of 293
lactic acid producing bacteria (e.g. probiotic culture powders) (data not published). We did 294
not find, as expected, an additional benefit in the use of this medium for the products 295
described here. 296
297
3.3. Growth in food matrices 298
The growth of both the pathogens and the background flora was determined in BPW and 299
BPW-S. PH in the enrichment was recorded. All food matrices were inoculated with low 300
levels of the target organism. 301
The target bacteria and the background flora reached approximately the same final 302
concentration after 24 h for PIF, milk powder, oat flakes and minced meat (Figure 3). For PIF 303
and oat flakes the levels of the background flora was 1 log cfu/ml lower in BPW-S than in 304
13
BPW throughout the incubation period. These are the products, which have a mainly Gram-305
positive flora. In PIF, the final number of Cronobacter was over 1 log higher in BPW-S. In 306
the other food matrices, there was no obvious difference between the enrichment media in the 307
growth of the selected pathogens. In PIF, oat flakes and minced meat the target organisms 308
reached their final value after 15 h. In milk powder it took 20 h until Salmonella reached the 309
maximum numbers. 310
Sprouts were spiked with a higher number of Salmonella to enable detection of the target 311
bacteria after the incubation period. It is remarkable that starting from 0 h of incubation, the 312
counts of the background flora are at least 3 log cfu/ml steps higher throughout the 313
enrichment period. 314
Background flora counts in sprouts were very high and values of 7 to 8 log cfu/g in sprouts 315
have already been published by other researches (Jinneman et al., 2012). In contrast to the 316
other matrices used in this study, the background flora of sprouts is mainly composed of 317
Gram-negative organisms and therefore the inhibiting agents in the BPW variants are not 318
effective because they were selected to supress the Gram-positive flora. Inhibiting effects of 319
the background flora on pathogens like E. coli O157 and Salmonella have been described 320
previously (Vold et al., 2000, Duffy et al., 1999). An approach explaining this behaviour is 321
the Jameson-effect. According to this model, two species compete for the resources in the 322
environment during their growth. Growth of both populations stops when the resources are 323
depleted or when metabolic products inhibiting growth have accumulated (Cornu et al., 324
2011). It is possible that this model also applies in sprouts. However, since growth in our 325
experiments was only determined up to 24 h, the growth behaviour of Salmonella after this 326
time is not known. 327
In the enrichment of dry food products, pH values remained relatively stable. Only in milk 328
powder did the pH in BPW-S decrease to 5.8. BPW with increased buffer concentration did 329
not completely prevent pH decline but usually kept the pH about 1 point higher. However, pH 330
14
decrease did in none of the products reach a level that would inhibit growth of the target 331
bacteria. 332
333
3.4. Detection limit in sprouts 334
Since the above-mentioned experiments indicated that low levels of Salmonella and STEC 335
spiked into sprouts could not be recovered, the inoculation level was increased and 336
enrichments were additionally tested with real-time PCR for the presence of the pathogens. 337
These results are summarized in Table 2. Salmonella could not be detected by real-time PCR, 338
when sprouts were inoculated with only 10 cfu. However, Salmonella counts were 104-10
5 339
cfu/ml, which indicates that there was growth of S. enterica during the enrichment. 340
At an inoculation level of 102 cfu, PCR results for Salmonella were only positive in 6xBPW. 341
At an inoculation level of 103
cfu, Salmonella was detected with PCR from BPW and 6xBPW 342
and cultural isolation yielded a minimal value of 1.0x105 cfu/ml. Nevertheless, levels of 4.0x 343
105 cfu/ml of Salmonella in BPW-S were not detected in the PCR, even though the test 344
protocol included an immunomagnetic separation step. Hence, at an inoculation level of 103
345
cfu (levels of ≥105 cfu/ml after the enrichment step), Salmonella could not be consistently 346
detected. This, however, seems not to be due to the different enrichment broths which were 347
evaluated. The high number of accompanying flora and the dilution steps which are applied in 348
the PCR (only small volumes are used) are the reasons for a high detection limit, which seems 349
to be between 105 and 10
6 cfu/ml after the enrichment step. 350
351
Badosa et al. (2009) have evaluated ISO enrichment real-time PCR methods with different 352
selective enrichment media and found that Salmonella was only detected from sprouts in 50% 353
of the cases when the inoculation level was 10 cfu in 25 g. 1 cfu was never detected even 354
though a selective enrichment was performed after the enrichment in BPW. Their method was 355
able to detect 1 cfu in 25 g of other fresh produce. Filtered soy sprout extracts did result in 356
15
positive detection for all inoculation levels, which suggest that the high levels of 357
accompanying flora caused the failure of the detection (Badosa et al., 2009). Taking the 358
present data into account, the possible presence of a low dose of Salmonella in 25 g of sprouts 359
cannot be reliably detected with a one-broth enrichment strategy. A selective enrichment is 360
necessary to promote selective growth of the target organism and to inhibit the background 361
flora. However, members of the common sprout flora (such as Citrobacter spp.) can also 362
grow in selective broths and can be misidentified as Salmonella on selective agar. 363
364
STEC were neither detected with PCR nor when using Rapid` E. coli 2 agar with the lower 365
inoculation levels. Isolation of the STEC strains on agar was only possible with an inoculation 366
level of 103 cfu. Still, the PCR result of the enrichment in BPW was negative. 367
Differences in cultural detection between STEC and Salmonella can be explained by different 368
selectivity of the agar media. The Salmonella strain was grown on LB agar supplemented 369
with nalidixic acid which supressed growth of any background flora on the plate. On the other 370
hand, Rapid` E. coli 2 agar does only select for coliforms. Since these organisms represent the 371
dominating flora of sprouts with counts of up to 105
cfu/g (Becker and Holzapfel, 1997), the 372
small number of E. coli present will be overgrown on the agar medium. The single 373
enrichment of sprouts in BPW followed by real-time PCR for the top 6 serotypes (O157, O26, 374
O145, O103, O111, O104) is the recommended ISO protocol for the detection of STEC 375
(ISO/TS 13136:2012). However, our results strongly suggest that low levels of contamination 376
with STEC on sprouts are missed with this procedure. Jinneman et al. (2012) found that the 377
method as described in the U.S. Bacteriological Analytical Manual did detect low levels of E. 378
coli O104 from a sprout matrix with different real-time PCR systems. Their enrichment 379
involved addition of acriflavin, cefsulodin and vancomycin and increased incubation 380
temperatures. However, they also reported non-isolation of the target organism on certain 381
selective agars due to overgrowth of background flora. 382
16
383
384
4. Conclusion 385
Different buffered peptone water based enrichment broths were tested to evaluate their 386
performance in a one-broth strategy for the detection of Gram-negative food borne pathogens. 387
Desiccation stressed cells showed variable behaviour. Resuscitation of Cronobacter spp. was 388
improved in BPW-S but stressed STEC cells did not show an improved recovery in pure 389
culture experiments. When a selection of food products was spiked with low numbers of the 390
pathogens, STEC and Salmonella growth was similar in all media. Cronobacter spp. detection 391
from PIF was improved with BPW-S. Therefore, we suggest the use of BPW-S in a one-broth 392
enrichment strategy for the isolation of Cronobacter spp. from PIF and similar products with 393
low aw values. Moreover, this broth can be used for the parallel enrichment of Cronobacter 394
spp., STEC and Salmonella from different food matrices. However, the one-broth strategy 395
was not suitable for produce with high numbers of Gram-negative accompanying microflora. 396
Low numbers of the pathogen will be overgrown by the background flora making their 397
detection impossible. Next should be a thorough validation of the medium according to ISO 398
standards. 399
400
401
402
References 403
Andrews, W.H., 1986. Resuscitation of injured Salmonella spp. and coliforms from foods. 404
Journal of Food Protection 49, 62-75. 405
406
Badosa, E., Chico, N., Pla, M., Pares, D., Montesinos, E., 2009. Evaluation of ISO 407
enrichment real-time PCR methods with internal amplification control and detection of 408
17
Listeria monocytogenes and Salmonella enterica in fresh fruit and vegetables. Letters in 409
Applied Microbiology 49, 105-111. 410
411
Baylis, C.L., MacPhee, S., Betts, R.P., 2000. Comparison of two commercial preparations of 412
buffered peptone water for the recovery and growth of Salmonella bacteria from foods. 413
Journal of Applied Microbiology 89, 501-510. 414
415
Becker, B., Holzapfel, W.H., 1997. Mikrobiologisches Risiko von fertigverpackten 416
Keimlingen und Maßnahmen zur Reduzierung ihrer mikrobiellen Belastung. Archiv für 417
Lebensmittelhygiene 48, 73-96. 418
419
Cornu, M., Billoir, E., Bergis, H., Beaufort, A., Zuliani, V., 2011. Modelling microbial 420
competition in food: application the behaviour of Listeria monocytogenes and lactic acid flora 421
in pork meat products. Food Microbiology 28, 639-647. 422
423
D`Mello, A., Yotis, W.W., 1987. The action of sodium deoxycholate on Escherichia coli. 424
Applied and Environmental Microbiology 53, 1994-6. 425
426
Duffy, G., Whiting, R.C., Sheridan, J.J., 1999. The effect of a competitive microflora, pH and 427
temperature on the growth kinetics of Escherichia coli O157:H7. Food Microbiology 16, 299-428
307. 429
430
Edel, W. and Kampelmacher, E.H., 1973. Comparative studies on the isolation of 431
“sublethally injured” Salmonellae in nine European laboratories. Bulletin of the World Health 432
Organization 48, 167-174. 433
434
18
Gray, V.L., Müller, C.T., Watkins, I.D., Lloyd, D., 2006. Peptones from diverse sources: 435
pivotal determinants of growth dynamics. Journal of Applied Microbiology 104, 554-565. 436
437
Hara-Kuo, Y., Ikedo, M., Kodaka, H., Nakagawa, H., Goto, K., Masuda, T., Konuma, H., 438
Kojima, T., Kumagai, S., 2000. Selective enrichment with a resuscitation step for isolation of 439
freeze injured Escherichia coli O157:H7 from foods. Applied and Environmental 440
Microbiology 66, 2866-2872. 441
442
ISO International Standard 6579, Microbiology of food and animal feeding stuffs- Horizontal 443
method for the detection of Salmonella spp. 4 th edition, 2002-07-15. 444
445
Jinneman, K., Waite-Cusic, Y.G., Yoshitomi, K. J., 2012. Evaluation of shiga-toxin 446
producing Escherichia coli (STEC) method for the detection and identification of STEC 447
O104 strains from sprouts. Food Microbiology 30, 321-328. 448
449
MacConkey, A.T., 1905. Lactose-fermenting bacteria in faeces. International Journal of 450
Hygiene and Environmental Health 5, 333-379. 451
452
MacDonald, I.C., Hackney, C.R., Ray, B., 1983. Enhanced recovery of injured Escherichia 453
coli by compounds that degrade hydrogen peroxide or block its formation. Applied and 454
Environmental Microbiology 45, 360-365. 455
456
Sata, S., Fujisawa, T., Osawa, R., Iguchi, A., Yamai, A., Shimada, T., 2003. An improved 457
enrichment broth for isolation of Escherichia coli O157, with specific reference to starved 458
cells, from radish sprouts. Applied and Environmental Microbiology 69, 1858-1860. 459
460
19
Stephens, P.J., Joynson, J.A., 1998. Direct inoculation into media containing bile salts and 461
antibiotics is unsuitable for the detection of acid/salt stressed Escherichia coli O157:H7. 462
Letters in Applied Microbiology 2, 147-151. 463
464
Stephens, P.J., Druggan, P., Nebe-von Caron, G., 2000. Stressed Salmonella are exposed to 465
reactive oxygen species from two independent sources during recovery in conventional 466
culture media. International Journal of Food Microbiology 60, 269-285. 467
468
Suh, J-H., Knabel, S.J., 2000. Comparison of different enrichment broths and background 469
flora for detection of heat-injured Listeria monocytogenes in whole milk. Journal of Food 470
Protection 64, 30-36. 471
472
Tamplin, M., 2002. Growth of Escherichia coli O157:H7 in raw ground beef stored at 10 473
degrees C and the influence of competitive bacterial flora, strain variation, and fat level. 474
Journal of Food Protection 65, 335-344. 475
476
Vold, L., Holck, A., Wasteson, Y., Nissen, H., 2000. High levels of background flora inhibits 477
growth of Escherichia coli O157:H7 in ground beef. International Journal of Food 478
Microbiology 56, 219-225. 479
480
Weber, C., Stephan, R., Druggan, P., Joosten, H., Iversen, C., 2009. Improving the 481
enrichment procedure for Enterobacteriaceae detection. Food Microbiology 26, 565-5 482
http://www.fda.gov/Food/FoodScienceResearch/LaboratoryMethods/ucm109656.htm 483
484
top related